Helium bubbling in a Molten Salt Fast

Reactor A flotation process

Dirkjan Journée Master of Science Thesis

Delft, Januari 2014

Supervisors: dr. ir. J.L. Kloosterman TNW, TU Delft

prof. dr. R. Konings TNW, TU Delft

Committee: prof. dr. R. Konings TNW, TU Delft

dr. ir. J.R. van Ommen TNW, TU Delft

dr. ir. M. Rohde TNW, TU Delft

Delft University of Technology Faculty of Applied Sciences

Dept. of Radiation Science & Technology

Nuclear Energy and Radiation Applications

2

3

Abstract

The Molten Salt Reactor (MSR) is one of the Generation IV nuclear reactor concepts. The molten salt

reactor uses a liquid fuel, which gives the possibility to reprocess the fuel while operating the reactor.

Reprocessing the fuel during operation strongly decreases the radiotoxicity of the waste and optimizes

the use of natural resources.

The first part of this thesis is to examine the reprocessing steps in a Molten Salt Reactor. There are

five reprocessing steps, which can be divided in on-line and off-line reprocessing. Helium bubbling is

used as the on-line reprocessing technique. Fluorination, protactinium removal, actinide extraction and

lanthanide extraction are used as the off-line reprocessing techniques.

It was found that the reprocessing steps show a promising molten salt reactor concept, but the

individual techniques are still to be optimized. And the different steps in the reprocessing of the salt

must still be connected together to form a continuous process. It was also found that the helium

bubbling technique, which was initially used to extract the gaseous fission products, also extracted a

part of the noble and semi-noble metals from the molten salt. These metals will not dissolve in the salt

and plate out on metal surfaces in the reactor. If the plate-out is excessive, the decay heat from the

noble and semi-noble metals may damage the reactor. The extraction of these metals with helium is

proposed to be achieved by a flotation process. The factors that play a role in a flotation process are

described by Schulze [Schulze, 1984] and Nguyen [Nguyen, 2004].

The second part and final goal of this thesis was to develop an experimental method to determine the

influence of the bubble size and gas flow in a flotation process. A flotation process is described as an

extraction of solids from a liquid with gas bubbles. The bubble size and gas flow are the variables that

are adjustable in the helium bubbling technique in a Molten Salt Reactor. The development of this

experimental method included the making of a new set-up with a newly fabricated piece of equipment

and developing a method to work with this set-up.

The result of the set-up is a modified version of a Hallimond tube. The set-up is built to use the

methods to examine a flotation process, which are described by Nguyen [Nguyen, 2004]. The methods

are a direct and an indirect. The indirect method is based on the extraction of a known amount of

particles that are added to the set-up. The direct method is filming the direct interaction of solids with

gas bubbles and is used to get a better understanding of how the process works. It was found that the

developed set-up is able to get good results for the direct method and promising results for the indirect

method. This new equipment and the new experimental method offer the possibility for a number of

follow-up studies on the topic of a flotation process.

4

5

Contents

Abstract ..................................................................................................................................... 3

Contents ..................................................................................................................................... 5

1 Introduction ........................................................................................................................... 7

1.1 Nuclear energy ................................................................................................................. 7

1.2 Generation IV reactors ................................................................................................... 10

1.3 Thorium fuel cycle ......................................................................................................... 11

1.4 Molten salt reactor .......................................................................................................... 12

2 Offline Reprocessing Steps in a Molten Salt Fast Reactor .............................................. 17

2.1 Fluorination .................................................................................................................... 17

2.2 Protactinium Removal .................................................................................................... 20

2.3 Actinide Extraction ........................................................................................................ 21

2.4 Lanthanide Extraction .................................................................................................... 25

2.5 Conclusion ...................................................................................................................... 27

3. Solid particles in a molten salt reactor ............................................................................. 29

3.1 Helium bubbling ............................................................................................................. 29

3.2 Materials in a flotation process ...................................................................................... 30

3.3 Flotation process theory ................................................................................................. 32

3.4 Conclusion ...................................................................................................................... 37

3.5 List of common symbols ................................................................................................ 38

4 Experimental simulation of the flotation process ............................................................. 39

4.1 Experimental .................................................................................................................. 40

4.2 Materials ......................................................................................................................... 46

4.3 Description of the experimental methods ....................................................................... 46

4.4 Results ............................................................................................................................ 48

4.5 Discussion ...................................................................................................................... 54

5 Conclusions .......................................................................................................................... 55

5.1 Literature study .............................................................................................................. 55

5.2 Set-up ............................................................................................................................. 55

5.3 Recommendations .......................................................................................................... 56

6

References ............................................................................................................................... 59

Appendix A ............................................................................................................................. 61

Reductive Extraction of Plutonium ...................................................................................... 61

Appendix B .............................................................................................................................. 63

Blueprint of the modified Hallimond tube ........................................................................... 63

Appendix C ............................................................................................................................. 65

Safety Assessment Sheet for the floatation set-up ............................................................... 65

Appendix D ............................................................................................................................. 69

Making a coordinate system in Paint.Net ............................................................................. 69

Appendix E .............................................................................................................................. 71

Determine the positions of the bubbles and particles on an image using Graph Grabber ... 71

Appendix F .............................................................................................................................. 73

Determine the volume of the particles and bubbles using Graph Grabber .......................... 73

Nomenclature .......................................................................................................................... 75

7

1 Introduction

The Molten Salt Reactor (MSR) is one of the Generation IV nuclear reactor concepts. The molten salt

reactor uses a liquid fuel, which gives the possibility to reprocess the fuel while operating the reactor.

Reprocessing the fuel during operation strongly decreases the radiotoxicity of the waste and optimizes

the use of natural resources. There are five reprocessing steps, which can be divided in on-line and off-

line reprocessing. Helium bubbling is used as the on-line reprocessing technique. Fluorination,

protactinium removal, actinide extraction and lanthanide extraction are used as the off-line

reprocessing techniques.

The first part of this study is to examine the reprocessing steps in a Molten Salt Reactor, where the

focus will be on the helium bubbling technique. Helium bubbling was initially used to extract the

gaseous fission products, but it was found that this technique also extracted a part of the noble and

semi-noble metals from the molten salt. These metals will not dissolve in the salt and plate out on

metal surfaces in the reactor. If the plate-out is excessive, the decay heat from the noble and semi-

noble metals may damage the reactor. Therefore it is important to extract these metals.

The extraction of these metals with helium is proposed to be achieved by a flotation process. The

second part of this research is to develop an experimental method to determine the influence of the

adjustable variables in a flotation process in the Molten Salt Reactor. The variables that can be

adjusted are the bubble size and the gas flow.

1.1 Nuclear energy

1.1.1 Worldwide Energy demand

According to the International Energy Outlook 2013 the worlds energy consumption will grow with an

estimated 56% between 2010 and 2040. Figure 1.1 shows a graph of the estimated energy

consumption growth in quadrillion BTU or Quad (1 quad = 293 TWh). Most of the increase in global

energy demand from 2010 to 2040 occurs among the developing nations outside the OECD

(Organization for Economic Cooperation and Development), particularly China.

Figure 1.1 World total energy consumption, 1990-2040 (in quadrillion Btu)

8

There are five different fuel types for energy production:

- Petroleum and liquid fuels1

- Coal

- Natural gas

- Renewables (e.g. hydropower)

- Nuclear

The two fastest growing energy sources are renewables and nuclear power [IEO2013, 2013]. The

impact of fossil fuel emissions on the environment and the sustained high world oil prices support the

expanded use of these sources.

1.1.2 Principle of a nuclear reactor

Nuclear power is generated by a nuclear reactor and converted into electrical energy. A nuclear reactor

is used to initiate and control a sustained nuclear fission reaction in order to generate electricity. A

nuclear reactor is part of the nuclear fuel cycle. The nuclear fuel cycle exists of the progression of

nuclear fuel through different steps. The steps can be divided into three different sections. The front

end, the service period and the back end. The front end are the preparation steps of the fuel, the service

period is the period where the fuel is used during reactor operation and the back end are the disposal or

reprocessing steps of the spent nuclear fuel. Figure 1.2 shows an overview of the nuclear fuel cycle of

uranium.

Figure 1.2. Overview of the nuclear fuel cycle of uranium.

The reactor converts the thermal energy released from nuclear fission reactions into electric energy.

The conversion from thermal energy to electricity is done by steam turbines. There are a lot of

different nuclear reactor designs, but the principle is: the energy from a nuclear fission reaction is used

to heat a working fluid (in most cases the working fluid is water or a gas) to generate steam which runs

through turbines to generate electricity. This principle design is shown in figure 1.3.

1 In IEO2013, the term petroleum and other liquid fuels includes a full array of liquid product supplies. Petroleum liquids include

crude oil and lease condensate, natural gas plant liquids, bitumen, extra-heavy oil, and refinery gains. Other liquids include gas-to-

liquids, coal-to-liquids, kerogen, and biofuels.

9

Figure 1.3. Principle of a nuclear reactor. The reactor core heats up a working fluid, in this case water, to evaporate

to steam. This steam runs through a turbine to generate electricity. The steam is condensed and pumped back to the

core.

The fission reactions are based on fissile materials that can sustain a chain reaction with neutrons.

When a fissile nucleus absorbs a neutron, it will fission into two lighter elements (fission products)

while releasing neutrons and a large amount of energy. The energy is released mainly due to the

kinetic energy of the fission products and the ionizing radiation. The free neutrons can be captured by

another fissile element to create a nuclear chain reaction. Figure 1.4 shows a drawing of a nuclear

chain reaction.

Figure 1.4. Nuclear chain reaction. A fissile nucleus captures a neutron to undergo fission while releasing energy and

free neutrons.

The released neutrons have an energy of 0.7 Mev - 10 MeV, these neutrons are called fast neutrons.

By means of a moderator the neutrons are slowed down till they have a velocity corresponding to the

most probable energy at 20 degrees Celsius, which is about 2.2 km/s, these neutrons are called thermal

neutrons and have an energy about 0.025 eV. Neutrons with different energies have different neutron

absorption cross-sections, which means that the probability of an interaction with a nucleus differs.

10

A nuclear reactor either uses a fast neutron spectrum or a thermal neutron spectrum. The different

fuels and moderators determine the neutron spectrum.

1.2 Generation IV reactors

There are a lot of different types of nuclear reactor designs. The different designs can be divided into

different groups, called ‘generations’. Generation I reactors are the early prototype reactors.

Generation II are the first commercial reactors. Generation III reactors are the more developed

Generation II reactors. The next generation reactors are the Generation IV reactors. The Generation IV

International Forum (GIF) considered six different type of reactors. These reactors have to measure up

to:

- reduced capital cost

- enhanced nuclear safety

- minimal generation of nuclear waste

- proliferation-resistance

The six different reactor types are summed up in table 1.1.

Table 1.1. Overview of Generation IV Reactors. [GIF, 2009]

System Neutron

Spectrum Fuel type

Outlet

Temperature ⁰C Size (MWe)

Very-high-temperature reactor

(VHTR) Thermal Solid 900 - 1000 250 - 300

Supercritical-water-cooled

reactor (SCWR) Thermal/Fast Solid 510 - 625

300 - 700

1000 - 1500

Gas-cooled fast reactor (GFR) Fast Solid 850 1200

Sodium-cooled fast reactor

(SFR) Fast Solid 550

30 - 150

300 - 1500

1000 - 2000

Lead-cooled fast reactor (LFR) Fast Solid 480 - 800

20 - 180

300 - 1200

600 - 1000

Molten salt reactor (MSR) Thermal/Fast Liquid 700 - 800 1000

The molten salt reactor is the only design to use liquid fuel. Not only the design of a reactor can

accomplish the criteria of the GIF, also the used fuel can have an influence. Uranium and plutonium

are traditionally used as a fuel for nuclear reactors, but in the 1960’s it was found that the

thorium/uranium-233 fuel cycle can also be used for a nuclear reactor. Despite the success, the MSR

program closed down in the early 1970’s in favor of the liquid metal fast-breeder reactor.

11

1.3 Thorium fuel cycle

Thorium is the again discovered fuel for nuclear energy. The interest in thorium as a nuclear fuel

comes mainly from:

- the abundance of thorium in the earth’s crust compared to uranium

- the lower levels of waste generation, especially transuranic elements like neptunium,

plutonium and americium

- the excellent non-proliferation credentials

Thorium itself is not fissile, but fertile. When a fissile fuel element captures a neutron it will undergo

fission, while releasing large amounts of energy. When a fertile element captures a neutron it will

become a fissile element.

Natural thorium is thorium-232. Thorium-232 is transmuted into fissile uranium-233 in order to use as

a nuclear fuel. The production of uranium-233 from thorium-232 can be seen as a cycle, which is

called the thorium fuel cycle (see figure 1.5). The thorium/uranium nuclear reactions are the

following:

1. Natural thorium-232 (Th-232) captures a neutron from a fission reaction and becomes

thorium-233 (Th-233).

2. Thorium-233 (Th-233) decays with a half life of 22 minutes, to protactinium-233 (Pa-233)

through β--decay.

3. Protactinium (Pa-233) decays in 27 days to uranium-233 through β--decay.

4. Uranium-233 (U-233) undergoes fission, while releasing energy and free neutrons. The

neutrons can be used to continue the cycle.

Figure 1.5. Thorium fuel cycle. Th-232 captures an neutron to Th-233. Th-233 decays to Pa-233 while releasing an

electron. Pa-233 decays to fissile U-233 while releasing an electron. U-233 absorbs an neutron to undergo fission

while releasing energy and free neutrons which can be captured by Th-232 to start the cycle again.

Ideally, protactinium has to be removed from the bulk to avoid the capture of a neutron. When

protactinium captures a neutron, it will become non-fissile uranium-234. The breeding of uranium-233

can best be done in a molten salt reactor. The possibility of selective extraction of actinides in a

12

molten salt reactor makes it possible to remove the protactinium from the bulk before it can capture a

neutron. This will be further explained in chapter 2.

1.4 Molten salt reactor

Nuclear reactors consist of fuel and a coolant. Most nuclear fuels contain heavy fissile elements that

are capable of nuclear fission. These are solid fuels, like UOX (Uranium diOXide) or MOX (Mixed

Oxide, MOX fuel consists of a mixture of plutonium and uranium), which are pressed in pellets. These

pellets are sealed in fuel rods. The MSR uses a liquid fuel, in contrast to the other Generation IV

reactors, which use solid fuel rods. This gives the reactor a couple of extra advantages and challenges.

The MSR operates at atmospheric pressure and a temperature in the range of 625 - 775 ⁰C. The molten

salt reactor uses a liquid fuel, which gives the reactor some advantages compared to the reactors with a

solid fuel [Goethem, 2008]:

- Optimum use of fissile materials

- Very low production of long-lived wastes

- Flexible waste management due to high burn-up and on-site reprocessing

- The fuel is automatically drained under accident conditions into passively cooled storage tanks

- Most gaseous fission products (krypton, xenon) are continuously removed

- No fuel fabrication is required

- Smaller equipment size because of the higher volumetric heat capacity of the salts

1.4.1 History of the molten salt reactor; ARE and MSRE

The history of the molten salt reactor starts in 1946 with the NEPA (Nuclear Energy for the Propulsion

of Aircraft) project. The NEPA project worked with the ARE (Aircraft Reactor Experiment) program,

which was an experiment to design a nuclear-powered bomber. The ARE was designed for operation

at temperatures in the region of 1400 ⁰C at a power of 1-3 MWth with a fluoride-salt fuel circulating

in a heterogeneous core. The moderator were hot-pressed BeO blocks cooled by circulating sodium

[Bettis, 1957]. The project was cancelled in 1961.

After the ARE the MSRE (Molten Salt Reactor Experiment) was started in the 1960s. The MSRE was

an experimental molten salt reactor at ORNL (Oak Ridge National Laboratory) which was constructed

in 1964 and went critical in 1965 until 1969. Figure 1.6 shows the MSRE.

13

Figure 1.6. Layout of the Molten Salt Reactor Experiment. 1. Reactor vessel, 2. Heat exchanger, 3. Fuel pump, 4.

Freeze flange, 5. Thermal shield, 6. Coolant pump, 7. Radiator, 8. Coolant drain tank, 9. Fans, 10. Fuel drain tank,

11. Flush tank, 12. Containment vessel, 13. Freeze valve.

The MSRE operated at 650 ⁰C at a power of 7.4 MWt. The reactor used molten LiF-BeF2-ZrF4-UF4

(65-29-5-1 mole%) as a fuel. First 235

UF4 (33% enrichment) was used. Later 233

UF4 was added to the

salt, making the MSRE the first reactor to be fuelled with this fissile material [Haubenreich, 1970].

The materials for the MSRE are shown in table 1.2.

Table 1.2. Materials used in the MSRE.

Fuel salt:

Composition

Properties at 650 ºC

Density

Specific heat

Thermal conductivity

Viscosity

Vapour pressure

Liquidus temperature

LiF-BeF2-ZrF4-UF4 (65.0 - 29.1 - 5.0 - 0.9 mole%)

2.3 g/cm3

2.0 .103 J/kg.ºC

1.43 W/m.ºC

0.008 kg/m.s

<1 .10-4

bar

434 ºC

Coolant salt LiF-BeF2 (66-34 mole%)

Moderator Grade CGB graphite

Salt containers Hastelloy-N (68-Ni, 17-Mo, 7-Cr, 5-Fe)

Cover gas Helium

During normal conditions the reactor was operating at a power of 8 MW. The inlet temperature of the

fuel salt in the reactor vessel was 632 ºC and the outlet temperature was 655 ºC. The fuel salt was

pumped around with a flow of 75.7 l/s. The inlet temperature of the coolant salt at the heat exchanger

was 546 ºC and the outlet temperature was 579 ºC. The flow of the coolant salt was 53.6 l/s. Helium

14

was used as a cover gas to remove gaseous fission products like xenon and krypton. The design flow

sheet is given in figure 1.7.

Figure 1.7. Design flow sheet of the MSRE. The reactor is operating at a power of 8 MW. The inlet temperature of

the fuel salt at the reactor vessel is 632 ºC (1170 ºF), the outlet temperature is 655 ºC (1210 ºF). The flow of the fuel

salt is 75.7 l/s (1200 GPM). The inlet temperature of the coolant salt at the heat exchanger is 546 ºC (1015 ºF) and the

outlet temperature is 579 ºC (1075 ºF). The flow of the coolant salt is 53.6 l/s (850 GPM).

The program was terminated in the early 1970s, in favour of the Liquid metal fast-breeder reactor

(LMFBR). Most of the knowledge for the newly designed molten salt reactor is based on the results

from the MSRE.

1.4.2 Molten Salt Fast Reactor

The interest for the molten salt reactor resumed after the design became one of the six Generation IV

reactors. There are different MSR-projects over the world. The European project is called the EVOL

project (Evaluation and Viability of Liquid Fuel Fast Reactor Systems). The participants are:

CNRS, France

Helmoltz-Zentrum Dresden-Rossendorf, Germany

INOPRO, France

Joint Research Center, European Commission, Belgium

Aubert et Duval, France

Karlsruher Institute für Technologie, Germany

Energovyzkum Spol S.R.O, Czech Republic

Technische Universiteit Delft, Netherlands

Politecnico Torino, Italy

Budapesti Muszaki es gazdasagtudomanyi egyete, Hungary

The project is developing on the Molten Salt Fast Reactor (MSFR) concept. A schematic concept of

the reactor is given in figure 1.8.

15

Figure 1.8. A schematic concept of the Molten Salt Fast Reactor

One big difference of the design of the MSFR with the MSRE is the absence of the moderator. The

neutron spectrum becomes fast, which will make it better suitable for the transmutation of long-lived

waste, while at the same time it is possible to breed 233

U from thorium. The breeding is done in a

fertile blanket, which is located around the core of the reactor, as shown in figure 1.9 [EVOL, 2012].

Figure 1.9. Simplified scheme of the MSFR system including the core, blanket and fuel heat exchangers.

The molten salt inside the fertile blankets is not in contact with the fuel salt and contains initially only

lithium fluoride and thorium fluoride, as shown in table 1.3. The fertile blankets are used for the

breeding of uranium from thorium.

16

Table 1.3. Composition and the properties of the salts used in a MSFR

Fuel salt

Fertile Blanket Salt

LiF-ThF4-233

UF4 (77.5 - 19.985 - 2.515 mole%)

Or

LiF-ThF4-(Pu-MA)F3 (77.5 - 16.068 - 6.432 mole%)

LiF-ThF4 (77.5 - 22.5 mole%)

Properties at 700 ºC for LiF-ThF4 (78-22 mol%)

Density

Specific heat capacity

Thermal conductivity

Viscosity

4.125 g/cm3

1.6 .103 J/kg.ºC

1.01 W/m.ºC

0.010 kg/m.s

One of the biggest advantages of the molten salt reactor is the reprocessing of the fuel during

deployment. A reprocessing scheme has been established to treat used fuel by extraction of fission

products. The extracted actinides are sent back into the reactor core. In this way, the wastes

radiotoxicity is strongly decreased and the use of natural resource is optimized. Figure 1.10 shows a

graph of the radiotoxicity of the waste of a Molten Salt Reactor (LFTR = Liquid Fluoride Thorium

Reactor) compared to a normal reactor (PWR = Pressurized Water Reactor).

Figure 1.10. Comparison of the radiotoxicity of two different reactor designs as a function of time.

The reprocessing units can be divided into online and offline reprocessing. Online reprocessing takes

place in the reactor core, where offline reprocessing takes place outside the reactor core. The next

chapter will go into further detail on the offline reprocessing.

17

2 Offline Reprocessing Steps in a Molten Salt Fast Reactor

The liquid fuel in a molten salt reactor makes it possible to reprocess the fuel, while the reactor is still

in operation. Reprocessing the salt decreases the wastes radiotoxicity, by extracting the radiotoxic

actinides from the used fuel and sending the actinides back into the reactor core.

The total reprocessing scheme consists of online and offline reprocessing. The online reprocessing is

done with a technique called helium bubbling. Helium bubbling is an online gaseous extraction with

helium bubbles to remove gaseous fission products, Xe and Kr, and noble and semi noble metals by a

flotation process. The offline reprocessing techniques are called:

- Fluorination

- Protactinium removal

- Actinide extraction

- Lanthanide extraction

The fuel can be tapped from the reactor to reprocess offline. Figure 2.1 shows a schematic drawing of

the different reprocessing steps.

Figure 2.1. Schematic drawing of the reprocessing steps of the fuel of a molten salt reactor. With online reprocessing

(1) Helium Bubbling, and offline reprocessing (2) Fluorination, (3) Protactinium Removal (4) Actinide Extraction

and (5) Lanthanide Extraction.

2.1 Fluorination

The fluorination step aims at removing the elements which form volatile fluorides, in general with

high oxidation states. The extraction is done with flame fluorination. The separation process is based

on specific properties of uranium, plutonium and neptunium. The current research and development

work in the area of Fluoride Volatility Method (FVM) is carried out at the semi-technological line

called FERDA. The line is shown in figure 2.2.

18

Figure 2.2. Process flow-sheet of Fluoride Volatility Method

Uranium, plutonium and neptunium form volatile hexafluorides (UF6, PuF6, NpF6), where the majority

of the fission products constitute non-volatile fluorides. The volatile fluorides have different boiling

points, therefore it is possible to separate them by fractional distillation. Table 2.1 shows the different

boiling temperatures and volatility of the different fluorides.

Table 2.1, Distribution of fluorinated spent fuel according to the volatility [Uhlíř, 2007]

Group I

(highly volatile)

Group II

(volatile)

Group III

(non-volatile)

Agent m.p. (ºC) b.p. (ºC) Agent m.p. (ºC) b.p. (ºC) Agent m.p. (ºC) b.p. (ºC)

Kr -157.2 -153.4 IF7 5 4 AmF4 Subl. 513

CF4 -184 -129 MoF6 17.6 33.9 RhF3 Subl. 600

Xe -111.8 -108.1 NpF6 54.8 55.2 SnF4 Subl. 705

TeF6 Subl. -38.6 TcF6 37.9 55.2 ZrF4 912 918

SeF6 Subl. 34.5 UF6 64 56.5 PuF4 1037 927

PuF6 51.9 62.2 CsF 703 1231

IF5 9.4 98 RbF 760 1410

SbF5 6 142.7 UF4 1036 1450

NbF5 80 235 AmF3 1427 2067

RuF5 101 280 CmF3 1406 2330

RuF6 51 70 YF3 1136 2230

RhF5 95.5 n/a BaF2 1353 2260

RhF6 70 73.5 EuF3 1276 2280

GdF3 1380 2280

CeF4 838 Decomp.

CeF3 1430 2330

PmF3 1410 2330

SmF3 1306 2330

SrF2 1400 2460

19

The following reactions will take place [Delpech, 2009];

Production of gaseous F2 and H2:

2 HF → F2 + H2 (2.1)

Fluorination of metal M:

MF(y-x) + x/2 F2(g) → MFy(g) (2.2)

e.g. UF4 + F2 → UF6 (2.3)

Hydrogenation:

MFy(g) + x/2 H2(g) → MF(y-x) (in the molten salt) + x HF (2.4)

e.g. UF6 (g) + H2 (g) → UF4 + 2 HF (2.5)

The separation of the actinides and the other fission products is done with a series of condensers,

which allow separating the higher and lower boiling point components [Tulackova, 2007].

To separate the uranium, neptunium and plutonium the sorption-desorption method is used.

Neptunium, plutonium and uranium fluorides can be adsorbed on a sodium fluoride bed (NaF), to

separate them from the other fluoridised fission products. Neptunium and uranium can easily be

desorbed with fluorine gas. Plutonium will form a thermally stable complex with the sodium fluoride

traps, PuF4·3NaF. For this reason plutonium has to be separated from the other elements before

entering the sodium fluoride traps. -It is possible to convert all plutonium to non-volatile PuF4, but a

part of the uranium will also be in the non-volatile form UF4.-

Separating uranium from neptunium can be done with magnesium fluoride traps (MgF2). UF6 is not

adsorbed on MgF2, while neptunium is partially adsorbed. Table 2.2 shows the separation efficiencies

of the different elements.

Table 2.2 Possible separation efficiencies of selected components by using fluorination. [Uhlíř, 2007]

Element Separation efficiency

(%) Method used after fluorination

Uranium 95 - 99.9 - Adsorbed on NaF

- Desorbed with F2-gas

Plutonium 98 - 99.5 - Adsorbed on NaF

- Forms complex; PuF4·3NaF

Neptunium 60 - 70

- Adsorbed on NaF

- Desorbed with F2-gas

- Adsorbed on MgF2

20

2.2 Protactinium Removal

It is important to remove the formed protactinium from the fertile blankets and the fuel salt, to

maximize the breeding of fissile uranium (233

U) from thorium (232

Th). The reaction scheme from

thorium to uranium is shown below:

232Th + n →

233Th (β-decay) →

233Pa (β-decay) →

233U (2.6)

When 232

Th captures an neutron it becomes 233

Th, which has a half-life of 22.2 min and undergoes β-

decay to 233

Pa. 233

Pa has a half-life of 27 days and undergoes β-decay to fissile 233

U.

When 233

Pa captures another neutron it becomes 234

Pa, with a half-life of 1.17 min. 234

Pa undergoes β-

decay to non-fissile 234

U:

233Pa + n →

234Pa (β-decay) →

234U (2.7)

The capture cross section of protactinium is 5 to 6 times higher than the capture cross section of

thorium, which makes it a big challenge to remove protactinium before it captures a new neutron, to

form non-fissile uranium. Figure 2.3 gives a visual representation of the different neutron absorption

cross sections. Table 2.3 gives the different cross sections in barns (10-24

cm2) for thermal neutrons.

Figure 2.3. Visual representation of the neutron absorption cross sections of thorium, protactinium, uranium and

plutonium for different neutron spectrums.

Table 2.3. Neutron absorption cross sections for different nuclei in a thermal neutron spectrum [Nucleonica]

Nuclide Capture Cross Section

(barns)

Fission Cross Section

(barns) 232

Th 7.40 - 233

Pa 40.0 - 233

U 45.3 531.4 235

U 98.8 584.4 238

U 2.7 - 239

Pu 270.8 748.0

21

Protactinium can be removed by ‘molten salt/liquid metal reductive extraction process’ or ‘metal

transfer process’, which is the same method used for the removal of the other present actinides. A

scheme of the isolation of protactinium is shown in figure 2.4.

Figure 2.4, Simplified scheme of protactinium isolation from MSR fuel [Uhlíř _2, 2007].

2.3 Actinide Extraction

The remaining actinides in the fuel salt are americium, curium and the fraction of remaining

neptunium and plutonium that isn’t removed at the fluorination step.

2.3.1 Reductive extraction of actinides in a bismuth-thorium pool

These actinides are now extracted from the molten salt by a reductive extraction. This extraction takes

place in a liquid metal solvent constituted of bismuth with metallic thorium or lithium [Delpech,

2009]. The thorium or lithium is the reductive agent, which will reduce the oxidised actinide. The

thorium or lithium in the bismuth is formed to ThF4 and LiF. The molten bismuth and molten salt are

in two immiscible liquid phases. Figure 2.5 shows a schematic drawing of the reductive extraction.

22

Figure 2.5. Schematic drawing of extraction and back extraction with Thorium as a reductive agent, where MS =

LiF-ThF4 and LM = molten Bismuth.

The equilibrium reactions that take place are:

AcFn (Salt) + n/4 Th(Bi) ↔ Ac(Bi) + n/4 ThF4 (Salt) (2.8)

Or

AcFn (Salt) + n Li(Bi) ↔ Ac(Bi) + n LiF(salt) (2.9)

In general form:

AcFy (Salt) + y/z R(Bi) ↔ Ac(Bi) + y/z RFz(salt) (2.10)

Ac = actinides

R = reductive agent

The back extraction is realized by anodic oxidation. In which the actinides oxidise and newly added

ThF4 reduces to Th. The half-reactions are:

At the cathode:

Ac(Bi) + 3 F- → AcF3 + 3e

- (2.11)

At the anode;

ThF4 + 4e- → Th + 4F

- (2.12)

23

With the equilibrium constant for equation (2.10):

(2.13)

( ) ( ) ( ) (2.14)

α(i) = activity of species i

x(i) = mole fraction of species i

γ(i) = activity coefficient of species i

If the distribution coefficients are defined as:

(2.15a)

And

(2.15b)

Equation (2.13) in combination with (2.15) in logarithmic form can be written as:

(

) (2.16)

The importance of the activity coefficient differs in the literature:

- According to J. Finne [Finne, 2005]:

“The values in a fluoride molten salt are generally ranges from 1 to 10 whereas

and

values vary between 10-14

and 10-6

. From relation (2.16) it can be seen that the outcome of

the extraction depends on the values of the activity coefficients of the different compounds to

be extracted both in the molten salt and in the liquid metal. Especially, the choice of the liquid

metal is of primary importance.”

- According to L.M. Ferris [Ferris, 1970]:

“If the components in the bismuth phase are present at low concentration (generally less than

0.002 atom fraction); values of and should be practically constant. If in the salt phase,

MFy was either a solute present at low concentration (less than 0.002 mole fraction) or was a

major component such as LiF, BeF2 or ThF4 whose concentration was not significantly

affected by the equilibrium involved. Consequently, activity coefficients for the components of

the salt phase, and would also be nearly constant. When the respective activity

coefficients are constant, equation (2.13) reduces to:

(2.17)

And equation (2.16) loses the activity term:

24

(2.18)”

2.3.2 Potential control of solvent metal

The extraction of actinides, without extracting lanthanides, is depending on the potential of the

reductive agent, thorium, which is given by [Chamelot, 2007]:

( )

( ) ( )

(2.19)

E0 = standard potential of redox system ThF4/Th (V)

R = gas constant (J/mol.K)

T = temperature (K)

F = constant of Faraday (C/mol)

α = activities in Molten Salt (ThF4) and Liquid Metal (Th) (-)

The efficiency of extraction can be calculated with [Delpech, 2008]:

( )

(2.20)

With,

( ( )

) (

( )

( )) (2.21)

n = number of moles in the liquid metal phase (LM) and the molten salt phase (MS).

The selectivity of one actinide (Ac(1)) over another actinide (Ac(2)) is given by:

( ( )

( ))

( ( ))

( ( )) ( ( ))

( ( ))

(2.22)

The amount of Thorium in the bismuth pool is adjustable, which also adjusts the potential of the

bismuth pool. Adjusting the potential will adjust the efficiency and selectivity of extracting different

nuclides. The redox potential for different actinides is shown in figure 2.6.

25

Figure 2.6. Redox potential of the different actinides and lanthanides [Delpech, 2008]

Appendix A shows an example of a selective extraction of plutonium from a sample containing

plutonium and neodymium.

There are more possibilities with the reductive extraction method:

- The choice for thorium or lithium is depended on the preferred selectivity and/or efficiency of

the extraction of a certain actinide. [Uhlir_2, 2007]

- The LiF–ThF4 and LiF–BeF2–ThF4 salt mixtures with comparatively high thorium

concentrations (≥20–22 mol %) could not be used for effective separation of lanthanides and

thorium in a system for extraction processing of fuel salts [Zagnit’ko, 2012].

- The method of reductive extraction can also be connected to the fertile blankets to extract the

protactinium to give the opportunity to breed uranium. This can be a process to continuously

extract newly formed protactinium.

2.4 Lanthanide Extraction

The extraction of the lanthanides is based on the same method as the extraction of the actinides. By

changing the composition of the metal (bismuth) bath provides a way to control the potential, in order

to selectively extract the lanthanides. See figure 2.7.

The anodic oxidation, the back-extraction of the lanthanides, will be executed with a molten salt

mixture of LiCl-KCl, to form LnClx [Auger, 2008]. The lanthanides will be oxidised with H2O vapour

to form Ln2O3, which will be treated as waste.

26

Figure 2.7. Schematic drawing of the lanthanide removal. Where Ms = molten salt (LiF-ThF4).

The reaction that takes place in the bismuth pool is:

4/3 LnF3 (Salt) + Th(Bi) ↔ 4/3 Ln(Bi) + ThF4 (Salt) (2.23)

The back extraction is realized by anodic oxidation. In which the lanthanides oxidise and ThF4 reduces

to Th. The half-reactions are:

At the cathode:

Ln(Bi) + 3 Cl- → LnCl3 + 3e

- (2.24)

LiCl + e- → Li + Cl

- (2.25)

The lanthanide oxidation precipitation occurs according to the following reactions:

2 LnCl3 + 3 H2O (g) ↔ 6 HCl(g) + Ln2O3 (2.26)

Li + HCl (g) ↔ LiCl + ½ H2 (g) (2.27)

The formed hydrogen gas is let into the bismuth pool to reduce the ThF4 excess. This excess occurred

through the reductive extraction of the lanthanides. The reaction between hydrogen and thorium

fluoride is:

ThF4 + 2 H2 ↔ Th(Bi) + 4 HF (2.28)

27

The final product of the lanthanides is Ln2O3 (equation 2.26), which will be treated as waste.

The lanthanides are neutron poisons, especially samarium. Figure 2.8 shows the neutron absorption

cross section of samarium compared to uranium and xenon.

Figure 2.8. Comparison of the neutron absorption cross section of xenon, samarium, uranium and plutonium. The

non-fissile samarium is one of the lanthanides with a big absorption cross-section, which makes it an important

element to remove from the fuel salt.

2.5 Conclusion

Although the reprocessing scheme of the molten salt fast reactor is still in a developing stage, it shows

a promising molten salt reactor concept. The individual techniques are still to be optimized. And the

different steps in the offline reprocessing of the salt must finally be connected together to form a

continuous process.

The insoluble fission products are the noble and semi noble metals. The online reprocessing or cleanup

technique, helium bubbling, has shown promising results in extracting these metals. Helium bubbling

is initially used to remove gaseous fission products, but will also remove a significant fraction of noble

and semi noble metals. The precise amounts are depended on salt circulation mode, circuit geometry,

flow speed, etc., and therefore difficult to predict. The next chapter will go into further detail on the

removal of solid particles using the online reprocessing technique, helium bubbling.

28

29

3. Solid particles in a molten salt reactor

Most of the fission products formed in the salt in a molten salt reactor form stable fluorides that

dissolve in the salt. Noble and semi-noble metals form multi-atom clusters in the molten salt and

ultimately plate-out on metal surfaces, like the heat exchangers. The noble metals produce significant

decay heat. If the plate-out is excessive, the decay heat from the noble metals may damage the heat

exchangers via overheating should a loss of cooling occur [Forsberg, 2006].

The metals are non-wetting and insoluble relative to the salt. Metals are nonpolar, where the salt is

polar. Therefore it is thought to remove the metals from the salt with a flotation process. An inert gas

will be bubbled into the reactor to attach to the metals floating in the salt and to avoid that the metals

plate out on the surface in the reactor. The metal particles will float with the bubbles to the surface of

the liquid by the buoyancy force of the bubbles. It was found that noble and semi-noble metals were

partly extracted from the salt with the online reprocessing technique of helium bubbling [Kedl, 1972].

This chapter focuses on the removal of the solid particles from the fuel salt using the helium bubbling

technique.

3.1 Helium bubbling

Online reprocessing of the salt is done with helium bubbling. Helium is led into the fuel pump, which

is located on top of the reactor, to extract gaseous fission products. This is a continuous process. Table

3.1 shows the elements which are volatile at a temperature of 630 ⁰C. The most important element that

will be extracted is the neutron poison xenon.

Table 3.1. List of elements with their boiling points. These elements are volatile at 630 ⁰C and will be removed from

the molten fuel salt, because of helium bubbling in the reactor

Name chemical

element

Symbol Boiling-point (⁰C)

Helium He -269

Hydrogen H -253

Fluorine F -188

Oxygen O -183

Krypton Kr -153

Xenon Xe -108

Radon Rn -62

Bromine Br 58.8

Iodine I 184

Phosphorus P 280

Astatine At 337

Mercury Hg 357

Sulfur S 445

Arsenic As 613

30

Helium bubbling extracts the gaseous fission product in 30 seconds [Merle, 2007]. Figure 3.1 shows

the neutron poisons still present in the molten salt after helium bubbling, based on their cross-sections.

Figure 3.1. Distribution of the fission products capture rate in the core of an molten salt reactor after helium

bubbling [Merle, 2005]

3.2 Materials in a flotation process

A part of the noble and semi-noble metals were found in the off-gas system of the MSRE. It is thought

that the migration of the noble metals is due to a flotation process between the metals, the molten salt

and helium bubbles. Flotation is a dynamic process resulting from interaction of forces acting in a

bubble-particle-fluid system. The inertia of the three phases plays an important role in the flotation

process. The different materials that are present in the flotation process in the molten salt reactor are

noble and semi-noble metals, the molten salt and helium gas.

A new experimental method was developed during this thesis project, to get a better understanding of

the different interactions in a flotation process. To avoid working with the very toxic fluorides, that are

present in the molten salt, there are different liquids and fluids considered, which should have

comparable properties as the materials in the molten salt reactor. The next sections are focussing on

choosing the right materials for experimental simulation of a flotation process in a molten salt reactor.

3.2.1 Solids

The solids in a molten salt reactor are noble and semi-noble metals, which form nonpolar multi-atom

clusters with high Palladium, Ruthenium, Rhodium and Technetium content. The densities of these

31

metals are given in table 3.2. The size of these clusters is unknown. The size of metals found in

conventional nuclear reactors is in the order of microns.

Table 3.2. Noble and semi-noble metals found in the molten salt reactor

Name Symbol Atomic Number Density (g/cm3)

Molybdenum Mo 42 10.28

Technetium Tc*

43 11.5*

Ruthenium Ru 44 12.45

Rhodium Rh 45 12.41

Palladium Pd 46 12.02 (* Tc has no stable isotope and will decay into another element, but with a halflife of 2.105 a)

The solids used to experimentally examine the influence in a flotation process are given in table 3.3

Table 3.3. Materials used to perform experimental research on a flotation process

Name Composes of Density (g/cm3) Diameter (µm) Polarity

Soda lime glass Metal Oxides 2.52 500 - 600 Polar

Polystyrene (C8H8)n 1.05 500 - 600 Non polar

3.2.2 Liquid

The liquid in the molten salt reactor is initially a molten salt composed of LiF-ThF4-233

UF4 (with 77,5

mol% LiF), which is a polar fluid. The exact characteristics for this salt are not known, but will be

similar to the known salt LiF-ThF4 (77.5 mol% LiF, 22.5 mol%ThF4) [EVOL, 2012]. Table 3.4 shows

the physicochemical properties of this salt.

Table 3.4: Physicochemical properties of LiF-ThF4 (77.5 mol% LiF, 22.5 mol%ThF4). Density [EVOL, 2012], dynamic

viscosity [EVOL, 2012], surface tension [Yajima, 1984] .

Name Density @ 700 ⁰C

(g/cm3)

Dynamic viscosity

(Pa.s)

Polarity Surface tension

(LiF-ThF4, 80-20 @ 722 ⁰C)

(N/m)

Molten Salt 4.1052 0.0072

polar 0.266 ± 5.4

Uranium tetrafluoride has a higher density than thorium tetrafluoride. The fuel salt will therefore have

an even higher density than is given in table 3.4.

The liquids used to experimentally examine the influence in a flotation process are given in table 3.5.

Table 3.5: Physicochemical properties of Isopar M and water.

Name Components Density

(g/cm3)

Dynamic

viscosity (Pa.s)

Polarity Surface

tension

(N/m)

Isopar M C11 – C16 0.788 0.0028 Non polar 0.0257

Water H2O 0.998 0.001 Polar 0.0718

32

3.2.3 Gas

Helium gas is used to remove gaseous fission products in the molten salt reactor. Helium is an inert

gas which doesn’t react with the particles or the salt. In the experiment air is used as a gas.

3.3 Flotation process theory

There are some factors that play an important role to predict the interactions between bubbles and

particles in a liquid. The following factors are used to predict the feasibility of a working flotation

process:

- Particle settling velocity

- Force balance on a particle at the fluid interface

- Bubble rise velocity

- Bubble-particle interaction efficiency

These factors are used to compare the properties of the materials, which are considered to use in

possible experimental simulation of the flotation process in a MSR.

3.3.1 Particle settling velocity

Due to gravity, particles tend to settle to the bottom of flotation cells. The velocity of the particle

settling is therefore required in modeling the bubble-particle interactions and the suspension of

particles in flotation.

When a particle is dropped in a liquid, there is a brief transient period as the particle experiences the

acceleration due to gravity. After the transient period, the particle falls with a constant terminal

velocity, which is termed the settling velocity (Vs).

The settling velocity is the velocity of the particle when the sum of de drag force (Fd) and the

buoyancy force on a particle (Fbp) equals the gravitational force (Fg) [Nguyen, 2004a]:

(3.1)

With,

(3.2a)

(3.2b)

(3.2c)

Solving equation (3.1) for Vs yields:

( ( )

)

(3.3)

33

Cd is the drag coefficient, which is a function of the particle Reynolds number, which is given by:

(3.4)

(3.5)

Vs can now be described as:

( )

(3.6)

Figure 3.2 shows the settling velocity for different materials depending on the particle diameter.

Figure 3.2, Settling velocity of a noble metal (Pd) in stationary molten salt (LiF-ThF4, 77.5-22.5 mol%) (NM),

polystyrene in water (PW) and glass particles (SiO2) in Isopar M (hydrocarbon) (GI).

The theoretical settling velocity of polystyrene in water is low compared to the other two. The density

of polystyrene is almost similar to water (1.05 g/cm3).

10 20 30 40 50 60 70 80 90 1000

0.002

0.004

0.006

0.008

0.01

0.012Settling velocity

Diameter (m)

Velo

city (

m/s

)

NM = 0.0029849 m/s

GI = 0.0016839 m/s

PW = 0.00016878 m/s

Noble metal - Molten Salt (NM)

Polystyrene - water (PW)

Glass particle - Isopar M (GI)

34

3.3.2 Force balance on a particle at the fluid interface

The forces in a bubble-particle-fluid system are described by Schulze [Schulze, 1984]. The force

balance for a single particle adhering to a gas bubble can be expressed as:

(3.7)

With,

Fg (the gravitational force) =

(3.7a)

Fs (the capillary pressure in the bubble) = ( )(

) (3.7b)

Fa (additional detaching force) =

(3.7c)

Fb (the static buoyancy force) =

(( ( ))

( ( ))) (3.7d)

Fh (hydrostatic pressure) = ( ) (3.7e)

Fc (the capillary force) = ( ) ( ) (3.7f)

A particle will stay at the gas-liquid interface, if the sum of the force balance is zero or positive. The

forces on a particle are shown in figure 3.3.

Figure 3.3. Forces on a solid particle in a gas-liquid system. [Schulze, 1983]

35

Figure 3.4. Sum of the forces on a solid particle in a gas-liquid system versus the angle ω for θ=45º and θ=90º. The

diameter of the particle is 50 µm and the z-value (see figure 3.2) is 10 µm.

From figure 3.4 it can be concluded that the maximum angle ω is independent of the type of solid or

fluid.

3.3.3 Bubble rise velocity

The bubble rise velocity is a function of the bubble geometry, the physical properties of the medium

and the physicochemical properties of the gas-liquid interface [Nguyen, 2004b].

The bubble rise velocity of small bubbles in contaminated fluid is given by:

{

( ) } (3.8)

Where UStokes is the bubble terminal velocity predicted by Stokes’ law and is described by:

0 20 40 60 80 100 120 140 160 180

-2

-1

0

1

2

3

4x 10

-5 Sum forces ( = 90)

angle ()

Forc

e (

N)

Noble metal - Molten Salt (NM)

Polystyrene - water (PW)

Glass particle - Isopar M (GI)

0 20 40 60 80 100 120 140 160 180

-2

-1

0

1

2

3

4x 10

-5 Sum forces ( = 45)

angle ()

Forc

e (

N)

Noble metal - Molten Salt (NM)

Polystyrene - water (PW)

Glass particle - Isopar M (GI)

36

(3.9)

The Archimedes number is defined by:

(3.10)

The terminal rise velocity for bigger bubbles is given by:

{

}

(3.11)

Where Mo is the Morton number, given by:

(3.12)

The constraints for a and b are given in table 3.4.

Table 3.4. Constraints for equations (3.8) and (3.11) expressed in terms of the Archimedes number and the Morton

number

Equation Constraints a b

(3.8) Ar ≤ 12332 - -

(3.11) 12332 ≤ Ar ≤ 3.158 Mo-0.46

1 0

(3.11) 3.158 Mo-0.46

≤ Ar ≤ 29.654 Mo-0.46

1.14 -0.176

(3.11) 29.654 Mo-0.46

≤ Ar ≤ 506.719 Mo-0.46

1.36 -0.280

(3.11) 506.719 Mo-0.46

≤ Ar 0.62 0

The difference in volume of a bubble will have a different rise velocity which gives a different shape

bubble. The different shapes of the bubbles will have an influence on the chance of colliding with a

particle. The bubble rise velocity in three different fluids is given in figure 3.5.

37

Figure 3.5. Bubble rise velocity versus the bubble diameter in a Molten Salt (77.5 mol% LiF, 22.5 mol%ThF4),

Isopar M and water.

3.3.4 Bubble-Particle Interaction Efficiency

The interaction between bubbles and particles is the key mechanism to the success of the flotation

process. The three major processes that play a role in the bubble-particle interaction are:

- Bubble-particle encounter process

- Bubble-particle attachment process

- Bubble-particle detachment process

The efficiencies of these processes are best described by the rate of the particles that interact with a

bubble, given in number of particles per unit of time. The general equation is:

(3.13)

Where,

j = c (collision), a(attachment) or d(detachment)

r = real rate

i = ideal rate

3.4 Conclusion

The helium bubbling technique, which was initially thought to be used to extract the gaseous fission

products from the molten salt reactor, can also be used to extract a significant fraction of the noble and

semi noble fission products from the fuel salt in the MSR. The migration of these metals is due to a

0 0.005 0.01 0.015 0.02 0.025 0.030

0.01

0.02

0.03

0.04

0.05

0.06

0.07Bubble rise velocity

Diameter bubble(m)

Velo

city b

ubble

(m

/s)

Molten salt

Isopar M

Water

38

flotation process. The factors that play a role in a flotation process are described by Schulze [Schulze,

1984] and Nguyen [Nguyen, 2004].

A new experimental method was developed during this thesis project, to verify the theories of Schulze

and Nguyen and to get a better understanding of a flotation process. The development of this method

includes the fabrication of a set-up and the development of a procedure to work with this set-up to gain

experimental results. The next chapter will go into further detail on the experimental simulations of the

flotation process.

3.5 List of common symbols

Symbol Description Units

a Numerical constant -

b Numerical constant -

bm Acceleration in the external field of flow m/s²

Cd Drag coefficient -

Ej Bubble-particle interaction efficiency -

Fa Additional detaching force kg.m/s²

Fb Static buoyancy force kg.m/s²

Fbp Buoyancy force on a particle kg.m/s²

Fc Capillary force kg.m/s²

Fd Drag force kg.m/s²

Fg Gravitational force kg.m/s²

Fh Hydrostatic pressure kg.m/s²

Fs Capillary pressure in bubble kg.m/s²

g Gravitational force m/s²

k Numerical constant (=0.95) -

Njr Rate of interacting particles 1/s

Nji Rate of interacting particles 1/s

Rb Bubble radius m

Rp Particle radius m

Vs Particle settling velocity m/s

U Terminal bubble rise velocity m/s

UStokes Terminal bubble rise velocity by Stokes’ law m/s

z0 Liquid height (figure 3.3) m

Greek symbols

μ Liquid viscosity kg/m.s

ρl Liquid density kg/m³

ρp Particle density kg/m³

σ Surface tension N/m

φ Angle at the three phase point (figure 3.3) -

ω Centre angle (figure 3.3) -

Dimensionless

groups

Ar Archimedes number

Mo Morton number

Re Reynolds number

39

4 Experimental simulation of the flotation process

In the flotation process in the molten salt reactor are two variables that can be controlled; the size of

the bubbles and the gas flow. The size of the bubbles can be adjusted by using different nozzles

through which the gas enters the reactor. The change of the gas flow also influences the size of the

bubbles.

There is a direct and an indirect method to find the influence of the bubble size and flow in the

flotation process of solid particles [Nguyen, 2004c]:

- In the indirect method, a known amount of particles is added to a liquid filled column. A

known volume of gas is bubbled through the liquid to encounter and attach to the particles.

The attached particles will finally detach from the bubbles when they reach the liquid surface

to be collected in a receiver in order to quantitatively measure the amount of attached

particles. An example of such a set-up is shown in figure 4.1. Figure 4.1 shows a modified

Hallimond tube, which is characterized by the slightly bended column with the concentrate

receiver.

- In the direct methods, the particle trajectories passing a rising bubble are photographed or

filmed. This method is used to find out how the bubbles interact with the particles and to get a

better understanding how the flotation process works.

Figure 4.1. Set-up with a modified Hallimond tube.

40

4.1 Experimental

4.1.1 Overview of the set-up

In order to meet the conditions for both methods it was found that the set-up could be best made out of

glass, because of the transparency, in order to make photos or movies from the process. Glass can also

be easily adjusted in shape and size by the glass manufacturer if it doesn’t meet up to the desired

requirements.

The set-up is placed in a ventilated box in order to suck off the vapours from the used liquids. A

schematic drawing is given in figure 4.2. A picture of the final result of the developed set-up is given

in figure 4.3. The equipment is further described in section 4.2. The descriptions of the used methods

are described in section 4.4.

Unit Description

C1 Glass column

C2 Modified Hallimond tube

C3 Glass separatory funnel

C4 Glass three-neck flask

E1 Main valve air supply

E2 Pressure controller

E3 Bronkhorst gas flow

controller

E4 Three-way valve

E5 Bubble distributor

S1 Septum

Figure 4.2. Schematic overview of the set-up for a flotation experiment, the black square represents the ventilated box in which the set-up is

situated.

Figure 4.3. A picture of the developed set-up.

41

4.1.2 Description of the used equipment

Column

The basis of set-up consists of a glass column (C1) with a length of 500 mm and an inner diameter of

26 mm. and the top of the column has a NS29 ground glass joint connection. Under the top of the

column is a tube connection with a diameter of 6mm. The bottom of the column has a KF25 flange

connection. Above the bottom are another tube connection with a diameter of 6 mm. and a septum.

The normal joints for these columns are flanges. For this set-up the top flange is replaced by a ground

glass joint to easily attach another glass piece. The tube connections are mounted on the column to be

able to fill and drain the column without disassembling the whole set-up. A septum is mounted on the

column to be able to let the gas enter the column through a needle. Figure 4.4 shows the top and

bottom of the column, with the adjustments.

Figure 4.4. Left: The top of the column with a NS29 ground glass joint and a tube connection. Right: The bottom of

the column with a septum on the left and a tube connection on the right. At the bottom is a KF 25 flange with a

clamp.

Hallimond tube

On top of the column is a specially made modified Hallimond tube (C2) attached. The ends of the

Hallimond tube consist of NS29 ground glass joints. The lower part of the Hallimond tube has a small

valve which can be opened and closed by hand. Under the valve is a NS14.5 ground glass joint. On top

of the Hallimond tube is a straight glass tube attached to the ground glass joint. The tube has a length

of 100 mm and a diameter of 26 mm.

The modified Hallimond tube is the most important piece of equipment of the set-up. The tube is

specially designed for this set-up in order to be able to perform a quantitative experiment. The

blueprint for this tube is given in Appendix B.

The tube is constructed with a curve of 140 degrees. The choice for a curve of 140 degrees is based on

earlier made Hallimond tubes. The angle of the curve is to be experimentally examined to find the

optimum angle in which the rising bubbles can rise through the curve without detaching the attached

particles due to a collision with the wall. On the edge of the tube is an extra reservoir to collect the

particles which fall down from the liquid-gas surface. At the bottom of the reservoir is a valve to be

able to close the system. Figure 4.5 shows the modified Hallimond tube.

42

Figure 4.5. The modified Hallimond tube connected to the top of the column. The figure on the right shows a drawing

of the left figure. The tube has two NS29 ground joints on the thicker parts (red clamps) and a NS14.5 ground joint

(yellow clamp) under the valve. The valve can be opened and closed by hand. The valve, as shown in the picture, is

open.

Three neck flask

Under the Hallimond tube is a three neck flask (C4) attached with two NS19 and one NS24 ground

glass joints. The NS14.5 glass joint from the Hallimond tube is attached to the NS19 joint of the three

neck flask with a transition piece NS14.5/19. The three neck flask has an NS24 tube connection on the

middle neck and a plug on the second NS19 joint.

A three-neck flask belongs to the common used laboratory equipment, which makes it easy to obtain.

The first neck of the flask is attached to the Hallimond tube. At the second neck a polystyrene tube is

attached to function as an opening to be able to completely fill the set-up with a fluid. The opening is

necessary to release the air inside an empty flask. At first it was thought to use a round flask with one

neck, which was connected to the Hallimond tube, but it was impossible to completely fill the set-up

with this flask. The third neck is closed with a plug. Figure 4.6 shows the three-neck flask in the set-

up.

Figure 4.6. A three-neck flask is connected to the Hallimond tube. The three-neck flask is used to collect the particles

that are floated to the surface by the gas flow. The first neck is connected to the Hallimond tube. A tube is connected

to the second neck to release the air inside the flask during the filling of the set-up. The third neck is sealed with a

plug.

43

Separatory funnel

The tube connection at the top of the column is attached to a separatory funnel with a polystyrene tube

with a diameter of 6mm.

It was found that a separatory funnel could best be used to get all the weighted particles in the column.

The separatory funnel has an outlet which is located at the bottom side. Above the output is a valve

which can be handled manually. It is possible to use the weight of the particles to add them to the

column, by simply opening the valve of the funnel, because the output of the funnel is at the bottom.

At first it was tried to use a small tube to add the particles in the column. But it was not possible to

release all the particles in the column. A peristaltic pump was also tested to get the particles in the

column. But the pump wasn’t be able to pump the particles evenly around. Figure 4.7 shows the

separatory funnel in the set-up.

Figure 4.7. The separatory funnel is used in the set-up to add the particles to the column. The valve can be opened

and closed by hand. The funnel is filled with a liquid and the weighted particles are added to the funnel, while the

valve is closed. When the particles are settled at the bottom the valve can be opened to add the particles to the

column.

Bubble distributor

A sintered stainless steel distributor plate (E5) is situated between a small container and the bottom of

the column. The distributor plate is a Sika R20. The small container is attached to column with a

flange.

The distributor plate is used to distribute the gas flow uniformly in the column. The container will

slowly be filled with the used liquid, which will enter through the pores of the distributor plate. The

gas enters the bubble distributor through a small tube with an opening at the bottom. Figure 4.8 shows

a picture of the distributor at the bottom of the column and a drawing of a cross section of the

container through which the gas enters the column.

44

Figure 4.8. The distributor plate is situated between the column and a small container. The small container has three

ends, from which two are closed and one is connected to the air supply. The container and the column are connected

with a flange clamp. The figure on the right shows a cross section of the bubble distributor.

Gas supply

The used gas for this experiment is air. The air in the set-up enters the ventilated box through the main

air valve. The air supply is then reduced from 6 bar gauge to 2 bar gauge by a pressure controller. The

pressure controller can be adjusted by hand. A pressure meter shows the outlet pressure. The pressure

controller is connected to a Bronkhorst gas flow controller. The Bronkhorst gas flow controller is

directed by a computer with the following software; FlowDDE -2nd

V4.67 and FlowView 7 V1.19.

FlowDDE is used to connect the controller to the computer and FlowView is used to set the flow to

the desired value. The output of this controller can be set between 0.00 and 1.00 l/min. It is also

possible to set FlowView to a maximum amount of gas to be added to the set-up.

From the gas flow controller, the air goes through a three-way valve. This valve can be set to direct the

air to the column via the bubble distributor or via a needle which can be inserted through the septum

on the side of the column.

All the tubing for the gas is made out of Teflon and has a diameter of 3mm. Figure 4.9 shows the

equipment used to direct the air to the column. Figure 4.10 shows the three-way valve and the bottom

of the column, where the air enters the set-up through the bubble distributor or the septum.

45

Figure 4.9. Left: Pressure controller. Right: Bronkhorst gas flow controller. The black arrows show the direction of

the gas flow. The Bronkhorst gas flow controller is controlled by a computer. The cable on top of the controller is the

power supply. The cable which is above the gas output is the data cable.

Figure 4.10. Left: Three-way valve. Right: Bottom of the column. The black arrows show the direction of the gas

flow. The three-way valve distributes the air to the bubble distributor or to the septum. The right picture shows also

the drain tap from the column.

Additional equipment

- An Olympus i-Speed 2 camera is used to film the process during the experiments. The camera

is connected to a computer display to see the live image in order to focus the camera. The

recording can be started or stopped with a second computer. This computer uses i-Speed

Software Suite software.

- FlowDDE -2nd

V4.67 and FlowView 7 V1.19 software

- A macro lens: TAMRON SP Di, AF 90mm 1:2.8, Macro 1:1, was used on the camera.

- Two types of distributor plates

- A sieve with pores of 75 μm is used to separate the solids from the liquid.

- Erlenmeyer flasks are used to store the liquids.

- Small ceramic containers to store the collected particles

46

- A stove is used to evaporate the liquid which is still on the particles. The particles will then be

weighed to calculate the collected amount.

4.2 Materials

The following materials are used in this experiment:

- Solid particles (polar or non polar)

- Fluid (polar or non polar)

- Air

The properties of the different solid particles are shown in table 4.1.

Table 4.1: Properties of the different solids

Name Components Density (g/cm3) Diameter (µm) Polarity

Soda lime glass Metal Oxides 2.52 500 Polar

Polystyrene (C8H8)n 1.05 500 Non polar

The properties of the different liquids are shown in table 4.2.

Table 4.2: Properties of the different liquids

Name Components Density

(g/cm3)

Dynamic

viscosity (Pa.s)

Polarity Surface

tension

(N/m) Isopar M C11 – C16 0.788 0.0028 Non polar 0.0257

Water H2O 0.998 0.001 Polar 0.0718

Ethanol C2H5OH 0.789 0.002 Polar 0.0224

The experiments are performed with:

- Soda lime glass particles with Isopar M, or

- Polystyrene particles with a water-ethanol mixture

4.3 Description of the experimental methods

A safety assessment sheet is needed, before starting the experiments. This sheet gives an overview of

the experiment that is being done. This sheet can be found in Appendix C.

Indirect method

The set-up was filled with a water-ethanol mixture with a known composition. A batch of polystyrene

particles is sieved in the range of 500 µm to 600 µm. And a weighed sample of 1.5 gram is added to

the separatory funnel. After the particles were settled on the valve of the separatory funnel, the valve

was opened. The particles entered the column at the top. The gas-flow controller was opened, when

the particles were homogeneously dispersed in the column. The controller is set to be opened until

0.01 l. of gas is added to the column, the controller stops the flow when this amount is reached. The

inlet-flow of the gas was set to 0.01, 0.02, 0.03, 0.04, 0.05 and 0.10 l/min., to compare the influence of

47

the flow on the attachment of the particles. This experiment is repeated with different compositions of

the liquid. The first experiment is done with an ethanol-water mixture with a density of 0.912 gr/cm3.

This corresponds to a mixture of 41 vol% ethanol. The second experiment is done with a mixture of

0.883 gr/cm3. This corresponds to a mixture of 55 vol% ethanol.

When a particle is attached to a gas bubble, it will flow with the bubble to the gas-liquid interface at

the top of the modified Hallimond tube, where the particle will detach from the bubble and stay at the

gas-liquid interface. The liquid-gas interface becomes saturated with particles in time. When a particle

reaches the saturated liquid-gas interface it will detach from the bubble and fall into the three-neck

flask. The column is slowly drained through the tap, after the flow of gas has stopped. The particles

that are at the gas-liquid interface are added to the collected particles in the three-neck flask, because

of the lowering of the liquid level due to the draining of the column. The three-neck flask is emptied

onto the sieve to collect the particles. The collected particles are transferred to a small ceramic

container, which is put into an oven for 30 minutes at 110 ºC, to evaporate the liquid on the particles.

The collected particles are weighed, after they leave the oven. The weighed samples are compared to

find the difference in the amount of collected particles per gas flow.

Direct method

The high-speed camera was added to the set-up to film the attachment between the particles and gas

bubbles, to get a better understanding of the flotation process. Figure 4.11 shows a picture of the

camera and the set-up.

Figure 4.11. A picture of a high speed camera which is filming the process. The red square indicates the display

which shows the live image of the camera.

48

4.4 Results

The indirect method is used to find a relation between different gas flows and the yield of the particles.

This method is also used to compare the yields for different densities of the liquid mixture of ethanol

and water. The direct method is used to find the mechanism of a flotation process.

4.4.1 Experimental results for the indirect method

First experiment: Ethanol - H2O mixture (41 - 59 vol%)

The column was filled with an ethanol - H2O mixture (41 - 59 vol%) with a density of 91.2 g/cm3. A

weighted amount of polystyrene particles (500 - 600 µm) is slowly added to the top of the column (see

table 5.1, Mass in). Air is added to the column, when the particles are homogeneously dispersed.

An amount of 0.010 litre of gas is added to the column. This experiment is repeated for different gas

flows. The results are shown in table 4.3 and figure 4.12.

Table 4.3. The yield of the collected particles for different gas flows. The entered volume of gas is 0.01 l. The density of

the liquid is 91.2 gr/cm3.The particles have a diameter in the range of 500 - 600 µm.

Flow (l/min) Mass in (gr) Mass out (gr) Yield (%)

0.01 1.498 1.179 78.7

0.02 1.522 1.313 86.3

0.03 1.539 1.433 93.1

0.04 1.500 1.375 91.7

0.05 1.495 1.452 97.1

0.10 1.537 1.483 96.5

Figure 4.12. Yield of the particles for different gas flows. The entered volume of gas is 0.01 l. The density of the liquid

is 91.2 gr/cm3.The particles have a diameter in the range of 500 - 600 µm.

The difference in yield when the starting amount of particles is increased is shown in table 4.4.

70

75

80

85

90

95

100

0 0.02 0.04 0.06 0.08 0.1 0.12

Yie

ld (

%)

Flow (l/min)

Yield vs Gas Flow

49

Table 4.4. The yield of the collected particles for a significant difference in the initial amount of the particles. The

entered volume of gas is 0.01 l. The density of the liquid is 91.2 g/cm3.The particles have a diameter in the range of 500

- 600 µm.

Flow (l/min) Mass in (g) Mass out (g) Yield (%)

0.01 1.498 1.179 78.7

0.01 4.515 3.977 88.1

Second experiment: Ethanol - H2O mixture (55 - 45 vol%)

The column was filled with an ethanol - H2O mixture (55 - 45 vol%) with a density of 88.3 g/cm3. A

weighted amount of polystyrene particles (500 - 600 µm) is slowly added to the top of the column. Air

is added to the column, when the particles are homogeneously dispersed.

An amount of 0.010 litre of gas is added to the column. This experiment is repeated for different gas

flows. The results are shown in table 4.5 and figure 4.13.

Table 4.5. The yield of the collected particles for different gas flows. The entered volume of gas is 0.01 l. The density of

the liquid is 88.3 gr/cm3.The particles have a diameter in the range of 500 - 600 µm.

Flow (l/min) Mass in (g) Mass out (g) Yield (%)

0.01 1.497 1.196 79.9

0.02 1.508 1.228 81.4

0.03 1.506 1.210 80.3

0.04 1.504 1.057 70.3

Figure 4.13. Yield of the particles for different gas flows. The entered volume of gas is 0.01 l. The density of the liquid

is 91.2 gr/cm3.The particles have a diameter in the range of 500 - 600 µm.

It was noticed that during the experiments with the current Hallimond tube, a small fraction of the

extracted particles flows back into the column to either interact with another bubble or flow down to

settle at the bottom of the column. There are two reasons for this behaviour. The first reason is the

collision of a bubble-particle with the wall of the curve which can lead to the detachment of a particle

70

75

80

85

90

95

100

0 0.01 0.02 0.03 0.04 0.05

Yie

ld (

%)

Flow (l/min)

Yield vs Gas Flow

50

from a bubble. The second reason is because of an underflow at the top of the column, which allows

the particles to flow over the collecting reservoir back into the column.

4.4.2 Experimental results for the direct method

4.4.2.1 Bubble size

The results for the direct method are stills from the movies made with a high speed camera. Figure

4.14 shows a series of stills of the bubbles and particles in a 55 vol% ethanol – H2O mixture, with

different gas flows.

Flow 0.01 l/min Flow 0.02 l/min

Flow 0.03 l/min Flow 0.04 l/min

Flow 0.10 l/min Flow 0.20 l/min

51

Flow 0.30 l/min Flow 0.50 l/min

Figure 4.14. A series of stills with different gas flows. The liquid is a mixture of ethanol and water (55 vol%

ethanol). The gas bubbles are air. The inner diameter of the column is 26 mm. The outer diameter is 30 mm.

4.4.2.2 Speed and trajectory of an attached particle

The attachment of particles is filmed with a high speed camera. With Paint.Net and Graph Grabber

V1.5.5 software it is possible to calculate the speed and follow the trajectory of the particles. Because

the diameter is known it is possible to make a coordinate system.

Figure 4.15 is an example of such a system, where two different floating particles and bubbles are

examined. The horizontal axes is made in paint with a straight line and copied vertically to get a line

with the same length. The black square indicates the same bubbles and particles, but in another

position. A description of the method to use Paint.Net to make a coordinate system is shown in

appendix D.

Figure 4.15. Two different particles with attached bubbles are floating upwards. A coordinate system is made with Paint.NET

software. The horizontal axis is copied and pasted vertically.

The positions of the particles and bubbles can be found and stored with Graph Grabber. The method to

use GraphGrabber is shown in appendix E. The coordinates are given in table 4.6 and plotted in figure

4.16.

52

Table 4.6. Found coordinates of floating bubbles and particles. The coordinates are found with graphgrabber

Object x-axis,

begin

y-axis, begin x-axis,

end

y-axis,

end

Difference

Δx;Δy

Trajectory

length

(mm)

Bubble 1 6.77 17.86 7.57 22.96 Δx 0.80

Δy 5.10

5.16

Particle 1 6.74 18.38 7.57 23.45 Δx 0.83

Δy 5.07

5.14

Bubble 2 9.85 17.86 10.68 20.82 Δx 0.82

Δy 2.95

3.07

Particle 2 9.91 18.33 10.70 21.28 Δx 0.79

Δy 2.96

3.06

Figure 4.16. The trajectory of floating bubbles with an attached particle. The trajectory is shown in a coordinate

system.

The pictures are saved under numbers. By comparing the numbers and using the frame rate of the

camera it is possible to calculate the speed of the bubbles with the attached particles. The distance

between the two positions is calculated and divided by the time to find the velocity of the bubble with

the particle. In this example:

- The frame rate is 200 frames per second.

- The first picture is number 930 and the second picture is number 1000. The time between the

two pictures is (1000-930)/200 seconds = 0.35 s.

The different speeds of the particles are shown in table 4.7.

Table 4.7. The calculated velocities of different floating particles.

Object Trajectory length

(mm)

Time

(s)

Velocity

(m/s)

Particle 1 5,14 0.35 0.0147

0.00

5.00

10.00

15.00

20.00

25.00

30.00

0.00 5.00 10.00 15.00 20.00 25.00 30.00

He

igth

of

the

co

lum

n

Width of the column

Trajectory of a rising bubble and a attached particle

Bubble/Particle 1 Bubble/Particle 2

53

Particle 2 3,06 0.35 0.0088

4.4.2.3 The volume of a bubble and particle

The volumes of the bubbles and attached particles are measured with Graph Grabber V1.5.5 software

the results are shown in figure 4.17. The bubbles and particles are the same as in figure 5.4.

Figure 4.17. Close-up of the bubbles and particles from figure 5.4.

The particles are considered to be spheres and the bubbles are considered to be spheroids. The

volumes of the different bubbles and particles are given in table 4.8.

Table 4.8. Volumes of the different bubbles and particles from figure 4.17, with the bubble/particle ratio and the

corresponding velocity of the bubble with the attached particle. The particles are considered to be spheres, the

bubbles are considered to be spheroids.

Object Volume

(mm3)

Ratio particle/bubble Velocity

(m/s)

Bubble 1 0.097 1.409 0.0147

Particle 1 0.137

Bubble 2 0.062 3.500 0.0088

Particle 2 0.218

The velocities of the two different couples can be compared to their particle/bubble ratio. This

comparison shows that the lower the ratio the higher the velocity (NB. this cannot yet be concluded to

be true for every situation). The method to use Graph Grabber to measure the volumes of the bubbles

and particles is described in Appendix F.

54

4.5 Discussion

The use of Isopar M and glass beads as materials for the experiments did not work properly, because

the Isopar M reacted with the polystyrene tubing. These experiments where therefore canceled.

4.5.1 Indirect method

It followed from the first experiment that the yield increases when the gas flow increases. The yield of

the extracted particles with a gas flow of 0.05 l/min (97.1 %) is nearly the same as with a gas flow of

0.10 l/min (96.5 %), which means that the yield reaches a maximum at a certain gas flow.

For the second experiment the same is true, although the yield was much lower. The maximum yield

for this experiment was 81.4 % and found at a gas flow of 0.02 l/min. The main difference of the two

experiments is the density difference of the liquid and the particles. The higher yields are found when

the density difference is low.

The yield increases when the “mass in” is increased. It was found in the direct method that one bubble

can extract multiple particles. So, when the concentration of particles increases the possibility of a

bubble colliding and interacting with multiple particles increases, which is a possible explanation for

this behaviour.

4.5.2 Direct method

The volume and the shape of the bubbles are depending on the gas flow. An increase of the gas flow

gives an increase of the bubble volume and eventually the shape. Until a gas flow of 0.20 l/min the

shape of the bubbles is mainly round. With higher gas flows the bubbles become more ellipse.

The speed of the bubble-particle is calculated by using different points of the trajectory of the bubble-

particle. This method is not yet perfect, because it is based on a 2 dimensional situation, while the

bubble-particles are in a 3 dimensional situation. The found velocities and trajectories are therefore not

completely accurate.

The ratio of the volume of the particle and the bubble are compared with the velocity of the bubble-

particle. This experiment shows that when the ratio becomes smaller, the velocity becomes higher.

This cannot yet be fully concluded on the comparison of just two bubble-particles and must therefore

be verified by multiple experiments on the ratio versus the velocity of bubble-particles.

55

5 Conclusions

The aim of this thesis was to first examine the different reprocessing steps in a Molten Salt Reactor.

And secondly, the investigation of the helium bubbling technique, which can also be used to extract

noble and semi-noble metals from the Molten Salt Reactor fuel.

5.1 Literature study

The first part of this thesis was to examine the reprocessing steps of the fuel salt in a Molten Salt

Reactor. There are five different reprocessing steps; helium bubbling, fluorination, protactinium

removal, actinide extraction and lanthanide extraction. These steps can be divided into online and

offline reprocessing. Helium bubbling technique belongs to the online reprocessing group and the

other reprocessing steps belong to the offline reprocessing group. The individual techniques are still to

be optimized. And the different steps in the offline reprocessing of the salt must finally be connected

together to form a continuous process.

The helium bubbling technique is used in the reactor to extract gaseous fission products from the fuel

salt on top of the reactor core. In the literature it was found that helium bubbling technique also

extracted noble and semi-noble metals from the fuel salt in this reactor. It is concluded that a flotation

process takes place during the helium bubbling reprocessing technique in the Molten Salt Reactor.

Therefore, there was an experimental method developed to determine the influence of the adjustable

variables in a flotation process in the Molten Salt Reactor. The variables that can be adjusted are the

bubble size and the gas flow.

5.2 Set-up

The second part and final goal of this thesis was to develop an experimental method to determine the

influence of the bubble size and gas flow in a flotation process. A flotation process is described as an

extraction of solids from a liquid with gas bubbles. The bubble size and gas flow are the variables that

are adjustable in the helium bubbling technique in a Molten Salt Reactor. The development of this

experimental method included the making of a new set-up with a newly fabricated piece of equipment

and developing a method to work with this set-up.

The final result of the set-up is a modified version of a Hallimond tube. The set-up is made out of

glass which has its advantages and disadvantages. One of the advantages of the glass parts of the set-

up is the transparency. The transparency makes it possible to follow the interactions in the process.

Another advantage is the possibility to easily adjust the shape of the glass. The major disadvantage of

the set-up made out of glass is that it is fragile; a small part of the glass column was broken during this

project when the set-up was disassembled for cleaning. Although broken parts can easily be restored,

it takes a lot of time, because this has to be done by an external company. The conclusion is that the

advantages are weighing up to the disadvantages.

The set-up is built to use the methods to examine a flotation process, which are described by Schulze

[Schulze, 1984] and Nguyen [Nguyen, 2004]. The methods are a direct and an indirect. The indirect

method is based on the extraction of a known amount of particles that are added to the set-up. The

direct method is filming the direct interaction of solids with gas bubbles and is used to get a better

understanding of how the process works. After testing the set-up it can be concluded that the

56

developed set-up is able to get good results for the direct method and promising results for the indirect

method. The indirect method has to be improved by altering the modified Hallimond tube. The

particles that initially were extracted by the bubbles are partially floating back into the column and

will be extracted again by the newly entered bubbles, which makes it impossible to say how much

bubbles are needed to extract a certain amount of particles. The set-up is therefore not yet optimised

for collecting quantitative results.

5.3 Recommendations

The recommendations for future research are divided into theoretical work, set-up improvements and

further research.

5.3.1 Theoretical work

To continue research on the topic of extracting noble and semi noble metals from the fuel salt of the

Molten Salt Reactor, it is recommended to examine the data of more bubble-particle interactions to

find a strong relation between the extractions of a particle in a known liquid by a bubble with known

volume and shape in a flotation process. The comparison of extraction yields from different flotation

processes should be used to predict the behaviour of the extraction of noble and semi-noble metals

from the fuel salt in the Molten Salt Reactor. The data that is needed to predict the behaviour of a

flotation process can be collected with the set-up described in this thesis.

5.3.2 Set-up improvements

The most important improvement for collecting quantitative results (indirect method) is adjusting the

modified Hallimond tube in such a way that the particles that are being extracted by the bubbles are all

collected. With the current tube, a small fraction of the extracted particles flows is not collected. There

are two reasons for this behaviour:

- The collision of a bubble-particle with the wall of the curve leads to the detachment of a

particle from a bubble.

- An underflow at the top of the column allows the particles to flow over the collecting reservoir

back into the column.

This behaviour can be prevented by adjusting the tube. There are a couple of options to prevent this

loss of extracted particles:

- The angle of the curve of the tube can be made larger to prevent the bubbles from colliding

too hard into the wall.

- By placing a glass barrier in the tube the particles can be prevented to float back into the

column.

- The diameter of the Hallimond tube can be widened to prevent the underflow to allow the

particles to flow back into the column.

The direct method is using the diameter of the glass to determine the size of the bubbles or the length

of the trajectory of the bubbles. The results don’t take into account the curvature of the glass. It has to

57

be examined what the deviation is of a particle or bubble in the glass per different liquid. It is therefore

recommended to use an object of known dimensions into the column to find the deviation on the film.

5.3.3 Further research

The set-up can be used to collect data for a better understanding of the mechanisms of a flotation

process. The methods that are described in this paper can be used to perform research on the

following:

- Examine the extraction yield of the particles with different liquid densities. This can also be

done with different type of liquids and particles.

- Find a relation between the extraction yield and the different gas flows in a flotation process.

The yield becomes maximum at a given gas flow, increasing the gas flow will than no longer

increase the yield. This knowledge can give a good inside in the optimal use of the gas supply.

The relation between the gas flow and the shape of the bubbles will probably play an

important role in this relation.

- Find a relation between the ratio of the volumes of the bubble-particles and the upward

velocity of the bubble-particles. This relation is dependent on the force balance between the

bubble and the attached particle, where the radius of the particles and bubbles play an

important role.

- Investigate the influence of different distributor plates on the volume and shape of the bubbles

at a given gas flow. This research can be combined with finding a relation between the density

of the used liquid and the volume and shape of the bubbles at a given gas flow.

- Examine the contact angles between the bubbles and the particles and their velocities. The

challenge in this research is to get very sharp close-ups of the interactions between bubbles

and particles.

58

59

References

[Auger, 2008] T. Auger, G. Barreau et al., The CNRS Research Program on the Thorium

cycle and the Molten Salt Reactors, Thorium Cycle – Molten Salt Reactors,

PaceN CNRS, 2008

[Bettis, 1957] E.S. Bettis, The Aircraft Reactor Experiment-Design and Construction,

Nuclear Science and Engineering, Volume 2, 804-825, 1957

[Chamelot, 2007] P. Chamelot, L. Massot et al., Feasibility of the electrochemical way in molten

fluorides for separating thorium and lanthanides and extracting lanthanides

from the solvent, Journal of Nuclear Materials 360, 64–74, 2007

[Delpech, 2008] S. Delpech, M. Salanne et al., Actinides/Lanthanides separation for the

thorium molten salt reactor fuel treatment, Atalante Montpellier, 2008

[Delpech, 2009] S. Delpech, E. Merle-Lucotte et al., Reactor physic and reprocessing scheme

for innovative molten salt, Journal of Fluorine Chemistry, Volume 130, 11-17,

2009

[EO2013, 2013] U.S. Energy Information Administration, International Energy Outlook 2013,

DOE/EIA-0484(2013), 2013

[EVOL, 2012] EVOL Project, Molten Salt Fast Reactor, Reference configuration, 2012

[Ferris, 1970] L.M. Ferris, Equilibrium distribution of actinide and lanthanide elements

between molten fluoride salts and liquid bismuth solutions, Journal of

Inorganic and Nuclear Chemistry 32, 2019 - 2035, 1970

[Finne, 2005] J. Finne, G. Picard et al., Molten salt/liquid metal extraction: Electrochemical

determination of activity coefficients in liquid metals, Journal of Nuclear

Materials 344, 165–168, 2005

[Forsberg, 2006] C. W. Forsberg, Molten-Salt-Reactor Technology Gaps, Proceedings of

ICAPP ’06, NV USA, paper 6295, 2006

[GIF, 2009] Gen IV International Forum, GIF R&D Outlook for Generation IV Nuclear

Energy Systems, 2009

[Goethem, 2008] G. Van Goethem, Generation IV reactor systems and fuel cycles (horizon

2030): technological breakthroughs in nuclear fission (int'l RD&DD), EC DG

RTD J2, 2008

[Haubenreich, 1970] P.N. Haubenreich, J.R. Engel, Experience with the Molten-Salt Reactor

Experiment, Nuclear Applications & Technology, Volume 8, 118-136, 1970

[Kedl, 1972] R. Kedl, The migration of fission products (noble metals) in the molten-salt

reactor experiment, ORNL-TM-3884, 1972

[Merle, 2005] E. Merle-Lucotte, L. Mathieu et al., Influence of the reprocessing on molten

salt reactor behavior, in2p3-00025413, version 1, 2005

60

[Merle, 2007] E. Merle-Lucotte, D. Heuer et al., The Thorium Molten Salt Reactor:

Launching the thorium cycle while losing the current fuel cycle, Université

Joseph Fourier, in2p3-00186944, version 1, 2007

[Nguyen, 2004] Anh V. Nguyen, Colloidal science of flotation, Surfactant Science Series

Volume 118, 2004

[Nguyen, 2004a] Anh V. Nguyen, Colloidal science of flotation, Surfactant Science Series

Volume 118, page 69 - 98, 2004

[Nguyen, 2004b] Anh V. Nguyen, Colloidal science of flotation, Surfactant Science Series

Volume 118, page 99 - 122, 2004

[Nguyen, 2004c] Anh V. Nguyen, Colloidal science of flotation, Surfactant Science Series

Volume 118, page 239 - 254, 2004

[Nucleonica] http://www.nucleonica.net/Applet/Decay/radioactive_decay.aspx

[Schulze, 1984] H.J Schulze, Physico-chemical elementary processes in flotation,

Developments in Mineral Processing, 1984

[Tulackova, 2007] R. Tulackova, K. Chuchvalcova-Bimova et al., Development of Pyrochemical

Reprocessing of the Spent Nuclear Fuel and Prospects of Closed Fuel Cycle,

Atom Indonesia, Volume 33, No. 1, 2007

[Uhlíř, 2007] J. Uhlíř, M. Mareček, M. Přeček, R&D on Fluoride Volatility Method for

Reprocessing of LWR and FR Oxide-type Fuels, Proceedings of ICAPP 2007,

Paper 7253, 2007

[Uhlíř _2, 2007] Jan Uhlir, Chemistry and technology of Molten Salt Reactors – history and

perspectives, Journal of Nuclear Materials 360, 6-11, 2007

[Yajima, 1984] K. Yajima, Surface Tensions of LiF-ThF4 Binary Melts and LiF-BeF2-ThF4

Ternary Melts, Journal of Chemical & Engineering Data, Volume 29, page

122-124, 1984

[Zagnit’ko, 2012] A. Zagnit’ko, V. Ignat’ev, Equilibrium Distribution of Lanthanum,

Neodymium, and Thorium between Fluoride Salt Melts and Liquid Bismuth,

Russian Journal of Physical Chemistry A, Vol. 86, No. 4, 533–538, 2012

61

Appendix A

Reductive Extraction of Plutonium

An example of the reductive extraction is done with the selective extraction of plutonium in a sample

of plutonium and neodymium, in a bismuth-pool with thorium. The selectivity equation is:

(

)

( )

( ) ( )

( ) (A.1)

An example of potential variation is given in figure 2.7. If the molar fraction of thorium in the bismuth

is equal to10-6

, and the potential of the liquid pool is -4.28 V, then, from figure 2.7, the extraction

efficiency of plutonium is of 99% and that of neodymium only 1%, according to the dotted line. For

lower concentrations of thorium and higher potential values, the extraction efficiency of plutonium

decreases. In this case, the extraction of large fraction of plutonium would require several extraction

stages [Delpech, 2009].

Figure A.1. Extraction efficiencies of a Bi-Th pool as a function of different potentials and thorium concentration in a

bismuth bath. The dotted line shows a potential of -4.28 V, which corresponds with an extraction efficiency of

plutonium of nearly 100% and for neodymium around 1%.

62

63

Appendix B

Blueprint of the modified Hallimond tube

Figure B.1. Blueprint of the modified Hallimond tube

64

65

Appendix C

Safety Assessment Sheet for the floatation set-up

Author: Name Dirkjan Journée Phone 0654694766

room nr. PF Rekenruimte

Area Supervisor: Name Ruud van Ommen Phone 015 27 82133

room nr. PF 02.[030]

Experiment

identification

Name Flotation tube with gas controller

location PF Ground floor Ventilated box

Description of experiment (including chemical equation, conditions and aim)

The goal of the experiments is to determine the effect of bubble size and flow in a three-phase

flotation process. A liquid (water/ethanol mixture) filled column is used with homogeneously

dispersed polystyrene particles (2 gram, 425 - 450 micron) and air is added as a gas.

The inlet gas flow is 6 barg, which is reduced by a pressure controller to 2 barg. A mass flow

controller controls the gas flow into column (max 1l/min).

Temperature (ºC) Room temperature Pressure (bar) Gas inlet (max): 6 barg

Column: Atmospheric

Labels that apply (chemicals and warning signs)

X X X

Emergency shut-down procedure

Close main valve of air supply and pressure controller

Remarks

Delft University of Technology

Faculty of Applied Sciences

Chemical Engineering department

ChemE

“Safety Assessment Sheet”

66

List of equipment:

1 glass column

1 Sintered stainless steel Distributor plate

1 modified Hallimond tube

1 three-neck flask

1 Separatory funnel

1 Bronkhorst gas flow controller

1 pressure controller

1 Three-way valve

1 computer

3 glass flasks

The setup is situated inside a ventilated box.

Assembly:

-Connect Bronkhorst gas flow controller to computer and close valve with the computer

-Connect gas inlet, pressure controller and gas flow controller

-Fix distributor plate

-Connect gas flow controller and distributor plate

-Fix column

-Fix Hallimond tube

-Fix three-neck flask

-Insert liquid

-Insert particles

Operation:

- Open main air valve and pressure controller

- Open valve of Bronkhorst gas flow controller

Shutdown, disassembly and cleaning

-Close main air valve and pressure controller

-Close valve of gas Bronkhorst gas flow controller

-Drain column via tap

-Detach three-neck flask

-Clean glasswork with water

Disposal of waste:

- Evaporate ethanol in open air inside the ventilated box

- Drain water through sink

- Dispose polystyrene beads as household waste

Approval of the area supervisor

Name Date: Valid through:

Signature

67

Chemicals (including gasses) and biological materials Chemical name and

related pictograms

Give H&P statements Storage

category

Polystyrene None given Solid

Glass Beads None given Solid

Ethanol

H225: Highly Flammable liquid and vapour

H302: Harmful if swallowed

H371: May cause damage to organs

P210: Keep away from heat/sparks/open flames/hot surfaces. No

smoking.

P260: Do not breathe dust/fume/gas/mist/vapours/spray.

Liquid

Isopar M

H304: May be fatal if swallowed and enters airways.

EUH066: Repeated exposure may cause skin dryness or cracking

P210: Keep away from heat/sparks/open flames/hot surfaces. No

smoking.

P301 + P310: If swallowed: Immediately call a POISON CENTER or

doctor/physician.

P331: Do NOT induce vomiting.

P370 + P378: In case of fire: Use water mist, foam, dry chemical

product of CO2 for extinction.

P403 + P235: Store in a well-ventilated place. Keep cool.

P405: Store locked up.

P501: Dispose of contents/container according to local rules.

Liquid

68

E1 Main valve air supply

E2 Pressure controller

E3 Bronkhorst gas flow controller

E4 Three-way valve

E5 Bubble distributor

C1 Glass column with modified Hallimond tube

C2 Separatory funnel

C3 Three-neck flask

69

Appendix D

Making a coordinate system in Paint.Net

Creating a coordinate system on an image is done with the following steps:

- open the image with Paint.Net (free software)

- make sure that there is free space around the image, e.g. the blank page in Paint.Net must

be bigger than the size of the image, in order to be able to work around the image [1]

- if the recorded image isn’t perpendicular, it is optional to tilt the image into the right angle

- add an extra layer to the image [2], the new layer is needed to be able to draw over the

image instead of on the image

- Make a straight horizontal line on this layer (button A), which will be the x-axis [3]

- Select the line and copy this line onto a third layer drag the line and rotate it (right mouse

button + Shift) into a vertical position, using the arrow keys makes it possible to move the

line with small steps into the right position, this will be the y-axis [4]

Figure D.1. An overview of the steps to make a coordinate system on an image.

70

71

Appendix E

Determine the positions of the bubbles and particles on an image using Graph

Grabber

Graph Grabber V1.5.5 is a free tool which is developed by Quintessa and can be found on:

- http://www.quintessa.org/software/downloads-and-demos/graph-grabber.html

GraphGrabber is a tool that enables you to retrieve data from graphs in the form of image files. It does

this by allowing users to import their graph image into the software and then to trace over it using

various graphing tools to capture the data.

The steps to collect the positions of certain particles or bubbles are:

- open an image (with known axis)

- use the buttons [1] to set the image in the right position

- use the next buttons [2] to set the axis

- with the button add data series, you can collect data from the image

- the data is exported from the file menu

Figure E.1. Screenshot of GraphGrabber

72

73

Appendix F

Determine the volume of the particles and bubbles using Graph Grabber

The volumes are calculated as follows:

- Particle =

o Where, d is the diameter of the particle

- Spheroid =

( )

o Where;

The different lengths are collected with GraphGrabber as shown in figure F.1.

Figure F.1. Using GraphGrabber to find the dimensions of the bubbles and particles

74

75

Nomenclature

Symbol Description Dimension

a Numerical constant -

b Numerical constant -

bm Acceleration in the external field of flow m/s²

Cd Drag coefficient -

D Distribution coefficient -

E0 Standard potential of redox system V

Ej Bubble-particle interaction efficiency -

Eff Extraction efficiency with actinide extraction -

F Faraday constant C/mol

g Gravitational force m/s²

k Numerical constant (=0.95) -

K Equilibrium constant -

Njr Rate of interacting particles 1/s

Nji Rate of interacting particles 1/s

R Gas constant J/mol.K

Rb Bubble radius m

Rp Particle radius m

S Selectivity at actinide extraction -

T Temperature K

Vs Particle settling velocity m/s

U Terminal bubble rise velocity m/s

UStokes Terminal bubble rise velocity by Stokes’ law m/s

z0 Liquid height (figure 3.3) m

Greek symbols

α(i) Activity of species i -

γ(i) Activity coefficient of species i -

μ Liquid viscosity kg/m.s

ρl Liquid density kg/m³

ρp Particle density kg/m³

σ Surface tension N/m

φ Angle at the three phase point (figure 3.3) -

χ(i) Mole fraction of species i -

ω Centre angle (figure 3.3) -

Dimensionless

groups

Ar Archimedes number

Mo Morton number

Re Reynolds number

Acronyms

Ac Actinides

ARE Aircraft Reactor Experiment

CNRS Centre National de la Recherche Scientifique

76

EVOL Evaluation and Viability of Liquid Fuel Fast

Reactor Systems

FVM Fluoride Volatility Method

GIF Generation IV International Forum

INOPRO Innovation-Optimization-Process

LFTR Liquid Fluoride Thorium Reactor

LM Liquid Metal

LMFBR Liquid Metal Fast-Breeder Reactor

MOX Mixed oxide

MS Molten Salt

MSR Molten Salt Reactor

MSFR Molten Salt Fast Reactor

MSRE Molten Salt Reactor Experiment

NEPA Nuclear Energy for the Propulsion of Aircraft

OECD Organization for Economic Cooperation and

Development

ORNL Oak Ridge National Laboratory

PWR Pressurized Water Reactor

R Reductive agent

UOX Uranium dioxide